What Actually Matters for the LinkedIn Algorithm

You'll hear that LinkedIn hides external links, punishes edited posts, or only promotes video. Some of that was true at some point. Much of it changes.

What stays consistent: the platform wants sessions where people read, comment, and come back tomorrow.

Your post gets a short test run with a subset of your connections and followers. If those people engage quickly and stick around, distribution expands. If they scroll past, it doesn't.

That model explains most of what you see in your analytics better than a new "algorithm update" thread every month.

Treat viral posts about the algorithm as entertainment, not ops. Your own last twenty posts are a better teacher than a screenshot of someone else's reach chart.

The first hour after you publish isn't everything, but it still counts.

Replies beat passive likes for signaling interest. Meaningful comments beat "Great post!" chains.

This is why a clear hook and a post that invites a specific response helps. Vague inspiration gets vague reactions.

You don't need to tag twenty people. You do need something worth responding to in the first few lines.

Posting when your audience is actually online gives that early window a fair shot. Timing isn't magic. It's basic logistics.

Reply to early comments with substance, not just thanks. Active threads keep the post warm while the platform decides whether to push it further.

LinkedIn measures whether people pause on your content, expand it, and read through.

Long posts aren't penalized by default. Boring long posts are.

Break up text for mobile. Use short paragraphs. Put the payoff early enough that readers don't bail in line three.

Carousels and documents can win here because swiping adds time on the post. Text can win too if the story earns each line.

If people like your post but nobody expands it, your hook or first paragraph is the bottleneck, not the algorithm.

Watch for the pattern across posts, not one bad day. One weak hook is a writing problem. Ten weak hooks is a process problem.

A post that travels far but reaches the wrong people fills your notifications with noise, not customers.

The feed tries to match content to interests inferred from what people engage with, who they follow, and what they post about.

Posting randomly across unrelated topics makes it harder for the platform to know who to show you to.

Founders do better when there's a through-line: the problem you solve, the buyers you talk to, the industry you operate in.

You don't need one niche forever. You do need enough consistency that strangers know why you're in their feed.

When reach drops, check whether your last five posts still look like the same person with the same point of view. Confused signals get confused distribution.

No single format always wins. Text, image, video, carousel: each can work for the right audience and topic.

What hurts performance is format mismatch. A dense essay pasted onto slide seven of a carousel. A video with no caption context. A link drop with no reason to click.

Native content generally gets a cleaner experience than posts that exist only to send people elsewhere. If you need a link, say what they'll get and make the post useful even if they don't click.

Experiment one format change at a time so you know what moved the numbers.

If you add video, caption it for silent scrollers. If you add a carousel, design slide one like a hook, not a title page.

Followers count slowly. Impressions swing post to post.

More useful: profile views from posts, comments from people you want to talk to, saves on educational content, and inbound messages that reference something you wrote.

Compare your last ten posts, not your last one. One viral outlier teaches less than a steady climb in reply quality.

If you're doing the basics (clear hook, consistent topics, readable structure) and still seeing flat reach, the issue is usually positioning or audience fit, not a secret setting you're missing.

Export analytics monthly. Circle three posts you'd write again and one you'd delete. That review beats refreshing impressions every hour.

Most "shadowban" stories are normal variance, a weak post, or an audience that's tired of the same angle.

True distribution collapse after months of steady performance is worth investigating: broken links, spam reports, or a sudden topic shift that confused your followers.

Before you assume punishment, publish three strong posts in your usual lane with better hooks. If reach recovers, you weren't banned. You were boring or inconsistent.

That's blunt, but it's usually the answer. The algorithm responds to reader behavior more than hidden switches.

If you're new to posting, low reach isn't a shadowban. It's a cold start. Keep the topics narrow and the hooks specific for six weeks before you diagnose platform problems.